The Role of Standardized Phase Angle in the Assessment of Nutritional Status and Clinical Outcomes in Cancer Patients: A Systematic Review of the Literature

Compared with the phase angle (PA), the predictive ability of the standardized phase angle (SPA) in assessing nutritional status and clinical outcomes in cancer patients remains uncertain. This review aimed to assess (1) the relationship between SPA and nutritional status and clinical outcomes (including complications and survival) in cancer patients; (2) the predictive ability of SPA alone and in comparison with the predictive ability of PA; and (3) the cut-off value of SPA in cancer patients. Studies that addressed the relationship of SPA use to nutritional status, complications, and survival in cancer patients were searched and identified from six electronic databases (PubMed, Medline, CINAHL, Embase, Web of Science, and the Cochrane Library). The included studies were considered to meet the following criteria: English studies with original data that reflected the effects of SPA on nutritional status and clinical outcomes (including complications and survival) and reported a cut-off value of SPA in cancer patients aged ≥18. Thirteen studies that included a total of 2787 participants were evaluated. Five studies assessed the relationship between SPA and nutritional status, and four of them reported a positive relationship between SPA and nutritional status in cancer patients, even considering SPA as a predictor. Twelve studies assessed the relationship between SPA and clinical outcomes in cancer patients. Two-thirds of the studies that evaluated complications reported the predictive ability of SPA; 30% of survival studies reported a positive relationship, 40% reported SPA as a predictor, and 30% reported no relationship. The standard cut-off value for SPA has not yet been determined. Data from the selected studies suggest that SPA might be a predictor of nutritional status. Further studies are needed to determine the value of SPA in predicting nutritional status and clinical outcomes in cancer patients.


Introduction
Cancer is recognized as a major disease worldwide. GLOBOCAN reported that the numbers of new cancer cases worldwide were approximately 18.1 million and 19.3 million in 2018 and 2020, respectively, and projected that in 2040 this number will be 28.4 million [1]. This indicates that the incidence of cancer is increasing year by year. Studies have shown that cancer patients often experience changes in body composition. The most common change is weight loss [2][3][4], and the loss of skeletal muscle causes a decrease in lean tissue [5,6]. At the same time, a variety of changes in adipose tissue have been shown to occur in different types of cancer [7][8][9]. Cancer patients often suffer from malnutrition [10][11][12] and poor clinical outcomes, including complications [13] and worse survival [14]. These factors can prolong hospitalization time [15], increase families' financial burdens [16], and even cause massive and disastrous expenditures by public health [17,18]. The European Society for Clinical Nutrition and Metabolism (ESPEN) reported that timely assessment of nutritional status can help patients maintain or gain weight and that this Table 1. Search strategy in electronic databases. Step Search Strategy #1 "Phase Angle" OR "Standardized Phase Angle"

Study Selection and Selection Criteria
The present study was designed and conducted according to PRISMA statements [42]. This review has been registered in the international prospective register of systematic reviews (PROSPERO) under registration number CRD42022327591. After obtaining the electronic database search results, two authors examined the titles, abstracts, and full texts of the studies. These two authors resolved disagreements at any review stage through negotiation and, if necessary, sought advice from a third person. Each original study was only included once. All types of studies were eligible, and there were no restrictions on sample size, sex, type of cancer, country, or region. Information on the BIA measurements was also not restricted by factors such as machine model or frequency.
The inclusion criteria were as follows: (1) the study participants were patients age ≥18 with cancer; (2) the study reported research on the effects of SPA on nutritional status and clinical outcomes (including complications and survival) in cancer patients; (3) a cut-off value for SPA was mentioned; (4) the article was published in English and included original data.
The exclusion criteria were as follows: (1) the participants were not cancer patients or were cancer patients aged ≤18; (2) SPA was not mentioned; (3) no effects of SPA on nutritional status and clinical outcomes (including complications and survival) in cancer patients, such as researching the screening effect of SPA on cancer or researching the effect of other intervention methods on SPA, were reported; (4) the study combined SPA with other indicators; (5) no cut-off SPA value was mentioned; (6) the article was a review, meta-analysis, meeting abstract, retraction, editorial, letter, personal comment, or book chapter, or did not present original data; (7) the article was not published in English.

Data Extraction
Two authors created a table that was used to extract information and extracted the key information in each study. The information that was recorded in the table included the first author, study title, year of publication, study country, main purpose of the research, study type, participants, sample size, cancer location, BIA machine model used, SPA cut-off value, and main findings. The number of participants in each study was based on the number of patients selected for inclusion in the study. The assessment tools were presented in an abbreviated form with the full name of each tool presented at the end of the form.

Quality Assessment
Two authors used the "Quantitative Non-randomized Studies" section of the Mixed Methods Appraisal Tool (MMAT) scale to assess the quality of the included studies. The MMAT is a reliable and valid tool for assessing the overall quality of the various study designs, which included five types: qualitative studies, quantitative randomized controlled trials, quantitative non-randomized studies, qualitative descriptive studies, and mixed method studies [43]. MMAT can be used to assess the quality of empirical studies, which are based on experimental work, and observational studies or simulations [43]. The specific items that appear in the MMAT are shown in Table S2; "*" means "meets the criteria", "-" means "does not achieve the criteria", and "/" means "we cannot find it in the study". Two authors marked the corresponding symbols in the corresponding table. Finally, we calculated the total scores of the included studies and assessed their quality levels.

Quality Assessment
Quality assessments of the included studies are presented in Supplementary Table  S2. All studies asked clear research questions that could be addressed by collecting data. All of the included studies were assessed using the "Quantitative Non-Randomized Studies" section of the MMAT. All of the participants in the 13 included studies were representative of the target population, all measures of outcomes were appropriate, and all exposures were as expected. One study had incomplete data due to the deaths of some of the cancer patients [47]; we marked this study as "-" in column 3 of Supplementary Table  S2. In addition, we marked two studies that did not consider possible confounders in their design and analysis [51,52] as "-" in column 4 of Supplementary Table S2. In conclusion, 10 of the included studies met 100% of the quality assessment criteria [13,34,40,[44][45][46][48][49][50]53], and 3 met 80% of the quality assessment criteria [47,51,52].

Quality Assessment
Quality assessments of the included studies are presented in Supplementary Table S2. All studies asked clear research questions that could be addressed by collecting data. All of the included studies were assessed using the "Quantitative Non-Randomized Studies" section of the MMAT. All of the participants in the 13 included studies were representative of the target population, all measures of outcomes were appropriate, and all exposures were as expected. One study had incomplete data due to the deaths of some of the cancer patients [47]; we marked this study as "-" in column 3 of Supplementary Table S2. In addition, we marked two studies that did not consider possible confounders in their design and analysis [51,52] as "-" in column 4 of Supplementary Table S2. In conclusion, 10 of the included studies met 100% of the quality assessment criteria [13,34,40,[44][45][46][48][49][50]53], and 3 met 80% of the quality assessment criteria [47,51,52].

Relationship between SPA and Nutrition Status
Five studies mentioned the relationship between SPA and nutritional status in cancer patients [13,34,47,49,53]; these are shown in Table 4 and Supplementary Table S3. In these studies, both subjective nutritional indicators (N = 2, 40%) [13,34] and objective nutritional indicators, including muscle function (N = 3, 60%) [13,51,53] and laboratory measurements (N = 2, 40%) [47,51], were mentioned. Pena et al. [13] assessed the relationship between SPA and nutritional status in 121 cancer patients who were awaiting surgery. They found that patients with SPA < −1.65 had decreased levels of PT-SGA, a parameter that has been listed as a recommended nutritional assessment tool for adults by the Australian Dietitians Association (DAA) [54]. Lower PT-SGA scores indicate worse nutritional status. In addition, arm circumference (MAC) [13,51], muscular midarm circumference (MMA) [13], calf circumference [51], handgrip strength (HGS) [13], triceps skinfold [13], and thigh adipose tissue [51], all of which are related to muscle function and form a part of the PT-SGA, also decreased with lower SPA. Poorer muscle function indicated poorer nutritional status. Similarly, Leon-Idougourram et al. [51] studied 45 patients with head and neck cancer who were undergoing systemic treatment; in that study, 26 patients with SPA < −1.65 had decreased levels of BMI (p = 0.04) and increased levels of C-reactive protein (CRP) (p = 0.04) and serum interleukin-6 (IL-6) (p = 0.007). Yates et al. [47] found a decrease in albumin levels (p = 0.014) when SPA < −1.65 in 100 patients with acute leukaemia. Although in recent years, ESPEN no longer recommends the use of serum albumin to identify adult malnutrition, and indicates that the decrease of serum albumin level is more indicative of the development of inflammation than malnutrition [55], ESPEN guidelines continue to recognize that inflammation is an important potential factor that increases the risk of malnutrition [55,56], suggesting that decreased serum albumin is associated with an increased risk of malnutrition in cancer patients. Therefore, these results of included studies directly or indirectly indicated that there was a higher risk of malnutrition in cancer patients who had lower SPA.
Based on the changes in SPA with nutritional status and related indicators, we considered the relationship between SPA and nutritional status and related indicators, and investigated whether SPA can be used as an indicator of nutritional status in cancer patients. As positive indicators of nutritional status, PT-SGA [13,34], HGS [13,53], MAC [13], MMA [13], BMI [53], and albumin (r = 0.20; p = 0.10) [47] were positively related to SPA. In contrast, weight loss [53], a negative indicator of nutritional status, was negatively related to SPA in 1084 cancer patients. Norman et al. [34] reported that SPA was an independent predictor of HGS (coefficient B 1.902; 95% CI, 1.321-2.483; p < 0.0001) and that SPA over the 5th percentile value had the strongest positive relationship to both moderate and severe malnutrition in 399 cancer patients, indicating that SPA was an independent predictor of nutritional status in cancer patients.

Relationship between SPA and Complications
Four studies evaluated the relationships between SPA and complications in cancer patients; the results of this evaluation are shown in Table 5.
Similarly, Harter et al. [52] found that patients with severe postoperative complications (including complications after surgical or radiological interventions or endoscopy, ICU, organ dysfunction, and even death) had a mean SPA value of −0.71 (−1.44; 0.16), while patients without complications had a mean SPA value of 0.41 (−0.16; 1.07), a significant difference (p = 0.007). They reported that patients with SPA < −1.65 had more complications than other cancer patients (p = 0.007). Roccamatisi et al. [40] found that SPA < 0.3 was associated with more complications (including multisite infections, infection with multidrug-resistant organisms, and candida coinfection) in patients with abdominal cancer (p = 0.032).
Regarding the relevance and predictive value of SPA in cancer patients, Pena et al. [13] reported a negative relationship between SPA and infectious complications in cancer patients (OR 4.19; 95% CI, 1.52-11.53; p = 0.006). SPA was considered as an independent predictor of infectious complications [13,40]. However, Pena et al. [13] also reported that there was no relationship between SPA and complications other than infectious ones (OR 1.07; 95% CI, 0.49-2.75; p = 0.881), and Maurício et al. [50] found that there was no significant relationship between SPA and postoperative complications in 84 cancer patients (RR 1.53; 95% CI, 0.79-2.92; p = 0.199).

Relationship between SPA and Survival
In addition to complications, survival is also an important indicator of clinical outcome in cancer patients. Urbain et al. [49] found that the survival rate of patients with SPA < −2.26 was 37% during a 2-year follow-up, while that of patients with SPA > −2.26 was 51%. In addition to the 2-year survival rate, the 5-year survival rate (p = 0.002) [45], the 60-day survival rate (OR 5.25; 95% CI, 1.35-20.44, p = 0.02), and the median OS (HR 1.57; 95% CI, 0.93-2.66; p = 0.09) [47] of cancer patients with lower SPA were also lower than those of patients with higher SPA. In addition, Paiva et al. [44] also found that the survival of cancer patients with SPA < −1.65 was on average 2 years less than that of cancer patients with SPA > −1.65 (p < 0.001).

Comparison of the Predictive Ability of SPA and PA
Five studies compared the prediction of nutritional status and clinical outcomes by SPA and PA in cancer patients; the results are shown in Table 7. Different studies have yielded different results. Norman et al. [34] reported that SPA adjusted for sex, age, and BMI enhanced the prognostic relationship of PA not only for nutritional status but also for clinical outcomes. Similarly, Paixao et al. [48] [40] also found that SPA (p = 0.032) predicted complications better than PA (p = 0.661) in cancer patients. However, Axelsson et al. [45] found that PA was a significant indicator for survival in cancer patients (HR 0.47, p < 0.001); it had an AUC of 0.73 in the ROC curve, higher than that of SPA (AUC 0.66). These results showed worse prediction of survival by SPA than by PA. Hui et al. [46] also found a stronger relationship between SPA (γ = 0.11; p = 0.11) than PA (γ = 0.075; p = 0.28) with clinicians' predictions of survival in cancer patients but a weaker relationship between SPA and nutritional-status-

Discussion
Although studies on predicting outcomes in cancer patients based on BIA have been increasing in number, the role of SPA has not been thoroughly studied. This review was designed to assess the usefulness of BIA-derived SPA in determining nutritional status and predicting clinical outcomes in cancer patients, and to perform a comparative analysis of the predictive efficacies of SPA and PA.
When malnutrition occurs, early membrane permeability increases, body fluid flows from intracellular water (ICW) to extracellular water (ECW), ECW/ICW increases and body cell mass decreases, adversely affecting the electrical properties of tissues, and PA is significantly reduced [57]. Well-nourished patients showed higher PA than malnourished patients [58][59][60][61]. A systematic review reported the predictive ability of PA for nutritional status in advanced cancer patients and found that low PA was related to worse nutritional status as assessed by BMI, serum albumin level, transferrin, and fat-free mass [62]. As a standardized form of PA, SPA has a great similarity to PA. Lower SPA indicated poorer nutritional status, especially malnutrition [63]. Therefore, this review sought to determine the role of SPA in assessing nutritional status in cancer patients. We found that SPA was related to nutritional status and that it showed lower values in HNC patients with elevated nutritional risk [51]; PT-SGA score, BMI and albumin level, all of which are recognized as vital assessment indicators of malnutrition, all showed a downward trend when malnutrition occurred [64,65], and these changes led to a decrease in SPA. Thus, SPA might play a role in predicting malnutrition in cancer patients. We found that SPA was positively related to PT-SGA [13,34] and BMI [53] in patients with mixed cancer and to albumin in patients with AL [47] but was negatively related to the degree of weight loss in patients with mixed cancer [53]. SPA was an independent predictor of nutritional status in cancer patients [34]. In patients with mixed cancer, malnutrition assessed by PT-SGA was also well predicted by SPA [34]. Norman et al. [34] reported that SPA's predictive ability was enhanced and better than that of PA based on assessment and quantification of individual deviations of cancer patients from population average levels for gender, age, and BMI. SPA offers practical advantages over conventional nutrition assessment methods in that it eliminates the need to measure weight and height when assessing nutritional risk.
Inflammation was identified as an important cause of malnutrition in the diagnostic consensus on malnutrition published by ESPEN [55]. Barrea et al. confirmed a negative relationship between PA and CRP, an inflammation-related factor, and reported the importance of PA in the diagnosis of meta-inflammation [66]. Our review also found that SPA is negatively related to levels of CRP and IL-6, two positive indicators of malnutrition [51]. Cancer cachexia arises from malnutrition [67] and is characterized by loss of muscle mass [68], which causes ECW to increase and ICW to decrease; thus, PA decreases [69]. The studies included in this meta-analysis showed that SPA was positively related to HGS [13,53], MAC [13], and MMA [13]. Based on these findings, SPA might reveal some changes in cancer patients' cachexia. Moreover, changes in impedance patterns (reduction of capacitive reactance and resistance retention) have been found to occur before overt symptoms of cachexia appear, suggesting a change in the electrical properties of tissues, especially somatic cell mass [70], and SPA decreased. Although patients with severe malnutrition are usually easily identified after screening, dietary assessment, or bedside examination, SPA offers a distinct advantage over current measurement methods for identifying patients without significant malnutrition; thus, SPA appears to be a good predictor of nutritional status in cancer patients.
Assessing the relationship between SPA and clinical outcome helps define the usefulness of SPA in cancer. A previous systematic review demonstrated that PA and SPA predict postoperative complications in cancer patients and encouraged greater reporting of SPA in future work [35]. This review found that, in cancer patients who were undergoing elective surgery (N = 60) or who had mixed [13] (N = 121) or abdominal cancer [40] (N = 182), SPA was significantly negatively related to and even the only predictor of complications. This might be related to the fact that during the occurrence and development of cancer, tumour-derived inflammatory cytokines are released, and homeostasis is damaged, increasing the risk of postoperative infectious complications and affecting the integrity of the cell membrane and somatic cell quality [13], and thus leading to reduced PA and SPA. However, one included study of colorectal cancer patients reported no significant relationship (N = 84) [50]. This might be due to heterogeneity among cancer patients. In addition, SPA and PA have been compared for their accuracy in predicting complications in cancer patients, and SPA was found to have better predictive ability [40]. Similarly, a meta-analysis reported that it was difficult to predict complications using PA in cancer patients due to differences in unadjusted factors such as age and sex, which could influence the interpretation of PA [35]. Therefore, after adjustment for confounders, SPA might be more effective in predicting complications in cancer patients.
Cancer patients often experience metabolic disorders and are prone to malnutrition or cachexia and other conditions that destroy the integrity of cell membranes and cause PA to decrease [71]. Currently, the work of several scholars has led to the use of methods that improve the survival of cancer patients by improving their metabolic status [72], and it was suggested that SPA might be of great significance in assessing survival. The investigation of the relationship and predictive ability of SPA regarding survival in cancer patients presented in this review showed that SPA was positively related to survival in patients with AL [47], patients with advanced cancer who received parenteral hydration treatment [43], and patients with mixed cancer [53]. SPA was also found to be a significant predictor of survival [34,44,45,49]. SPA was a good predictor of 6-month survival in patients with mixed cancers, and its predictive ability was improved compared to PA stratified by sex, age, and BMI [34]. In patients with haematological malignancies, SPA below the 25th percentile was a significant independent predictor of 2-year survival [49]. A stronger prediction of clinical outcomes by SPA than by PA in cancer patients was reported in the included studies [34,40,46,48]. However, the prediction of SPA for long-term survival in patients with stage 1 and 2 mixed cancer was reduced compared with the prediction for short-term survival in patients with stage 3 and 4 cancer [44]. A similar finding was reported for SPA prediction of 5-year survival in patients with HNC; although SPA was adjusted for PA, its prediction was lower than that of PA [45], possibly because SPA, as a prognostic tool, was very sensitive. It corrected for two important negative indicators, increased age and decreased BMI, and increased accuracy while reducing predictability. A 2021 meta-analysis reported that PA was an independent prognostic indicator of survival in patients with advanced cancer after adjustment for any possible confounding factors by multivariate Cox regression analysis [73]. The authors of that study pointed out that SPA could be influenced by adjusting PA for patients with different ethnicities and that this might cause inaccurate predictions when using SPA [73]. In addition, the prediction would change with the passage of time, leading to a decrease in the prediction of long-term survival. Notably, in two different studies, it was also shown that SPA was not related either to survival in patients with mixed cancer [13] or those who were undergoing radiotherapy [48], or to 30-or 60-day survival in patients with AL who were undergoing chemotherapy [47]. The reason for the different effectiveness of SPA in predicting survival might be that SPA was also influenced by the treatment the patients received, in addition to the tumour itself. Chemotherapy affects cell membrane function, calcium channels, and growth receptors [74]. Similarly, radiation has been shown to damage the integrity of cell membranes and increase their permeability [75]. All of these changes could lead to a reduction in SPA. In addition, as a common means of cancer treatment, radiotherapy and chemotherapy can prolong the survival of patients to a certain extent. These factors made it difficult to explore the relationship between SPA and survival rate.
In the clinical setting, determining an appropriate cut-off value for SPA will enable more accurate assessment based on testing nutritional status and clinical outcomes in cancer patients. A previous meta-analysis reported no uniform critical value for PA due to individual differences [76], and a 2021 meta-analysis indicated that the cut-off values of PA ranged from 4.73 • to 6 • in cancer patients, with a single cut-off value of PA yet to be determined [61]. Although individual bias was accounted for, there was still no uniform cut-off value for SPA in the assessment of nutritional status and clinical outcomes in cancer patients. The cut-off values for SPA in this review varied from −2.26 to 0.3, and the most commonly used cut-off values were −1.65 [13,40,44,45,48,[50][51][52] and the 5th percentile [34,46]. However, some studies used the 25th percentile because the Akaike information criterion (AIC) at the 25th percentile (AIC = 529.37) was lower than that at -1.65 (AIC = 531.48) and the 5th percentile (AIC = 530.80) [47,49], suggesting that the assessment of SPA at the 25th percentile was more concise and accurate. Technical factors were also one of the reasons for diversification of the SPA values. As there was no unified international standard, measurement differences in BIA arising from the use of equipment produced by different manufacturers [77] and nonspecification of electroneutral contact electrodes [78] affected the effective measurement of SPA.
Some limitations of this study must be acknowledged. The number of studies that assessed SPA as the main outcome was relatively low (there were only two such multi-centre studies). Due to the different types of BIA instruments used, there was no uniform standard for assessing SPA. All of these factors increased the risk of bias.

Conclusions
In conclusion, SPA has become an effective objective indicator for assessing the nutritional status of cancer patients. A lower SPA indicates a higher risk of malnutrition and provides a reliable basis on which more appropriate diagnostic and treatment methods can be chosen for clinical cancer patients. However, we still cannot draw definite conclusions about the ability of SPA to predict clinical outcome (including complications and survival rate) in cancer patients. No standard SPA cut-off value has been established. At present, hard evidence is still lacking but, given some promising research results, this may have been caused by the small amount of literature we included. Therefore, additional highquality studies are needed in the future, and more studies are necessary to clearly prove the relationship between SPA and clinical outcomes of cancer patients, and to establish the critical value of SPA. This will be of great significance in clinical diagnosis, management of treatment, nursing work, and the work of other health professionals in the future.
Supplementary Materials: The following supporting information can be downloaded at: https: //www.mdpi.com/article/10.3390/nu15010050/s1, Table S1: Search strategy details performed at 17 April 2022. Table S2: Quality assessments of included studies using by MMAT. Table S3: Mentioned outcomes related to SPA. Supplementary data to this article can be found in the paperwork we filed together.